Serum folate and vitamin B12 concentrations in relation to prostate cancer risk-a Norwegian population-based nested case-control study of 3000 cases and 3000 controls within the JANUS cohort

Background Although individual studies have been inconsistent, meta-analyses of epidemiological data suggest that high folate and vitamin B12 levels may be associated with increased prostate cancer risk. Methods Within JANUS, a prospective cohort in Norway (n = 317 000) with baseline serum samples, we conducted a nested case-control study among 3000 prostate cancer cases and 3000 controls, matched on age and time at serum sampling, and county of residence. Using conditional logistic regression, odds ratios (OR) and 95% confidence intervals (CI) for prostate cancer risk were estimated according to quintiles of serum folate, vitamin B12, methylmalonic acid (MMA), total homocysteine (tHcy) and methionine, and according to MTHFR 677C -> T genotypes. To correct for degradation during sample storage, folate concentration was measured as p-aminobenzoylglutamate (pABG) equivalents following oxidation and acid hydrolysis. Results We observed a weak positive association between folate concentration and prostate cancer risk [OR highest vs lowest quintile = 1.15 (0.97-1.37), P-trend = 0.04], which was more pronounced among individuals epsilon 50 years at inclusion [OR 1.40 (1.07-1.84), P-trend = 0.02]. tHcy showed an inverse trend with risk [OR 0.92 (0.77-1.10), P-trend = 0.03]. Vitamin B12, MMA and methionine concentrations were not associated with prostate cancer risk. Compared with the MTHFR 677CC genotype, the CT and TT variants, both of which were related to lower folate concentrations, were associated with reduced prostate cancer risk [OR 0.82 (0.72-0.94) and OR 0.78 (0.64-0.94), respectively]. Conclusion This large-scale population-based study suggests that high serum folate concentration may be associated with modestly increased prostate cancer risk. We did not observe an association between vitamin B12 status and prostate cancer risk.